Underage alcohol use remains at unacceptably high levels and is an enormous public health issue with a wide range of serious physical, social and psycho-social consequences, both acute and long-term. The goal of this application is to examine the extent to which underage alcohol use clusters geographically in a sample of US neighborhoods and cities and to elucidate a possible explanation for this clustering as it relates to individual, social and community-level environmental contexts. Despite the proliferation of efforts to change the community-wide environment surrounding alcohol use, estimates of the clustering of underage alcohol use are sparse and the manner in which contextual factors are related to this clustering have not been adequately examined. Clustering is most often treated as a "design effect" in epidemiologic surveys, but clustering can have important substantive implications. This application proposes secondary data analysis of the largest-ever randomized community trial of underage alcohol use to: (1) provide the first estimates of the magnitude of clustering of underage alcohol use based on pair wise odds ratios obtained from a statistical method called alternating logistic regression (ALR), (2) evaluate the role of individual, social and community-level environmental contexts on this clustering, and (3) develop a statistical approach for examining the clustering of subtypes of underage drinkers derived from latent class analysis. The proposed study will capitalize on data from repeated cross-sectional samples of approximately 6800 youth aged 14-20 from 68 US cities surveyed in each of three years (2004, 2006 and 2007) and collected as part of the Enforcing Underage Drinking Laws Randomized Community Trial (EUDL-CT). In addition to measures of alcohol use and alcohol-related problems, measures will include a wide range of contextual factors expected to relate to the clustering of underage alcohol use: social influences (friends, peers, family, and school), social context of drinking, neighborhood composition, community-level law enforcement activities, and alcohol outlet density. To the extent to which we can understand what causes clustering of underage alcohol use, we may identify aspects of the environment that are shared within communities and that may help to account for the higher or lower prevalence of underage alcohol use. This information will have importance by disclosing what contextual factors should be targeted in future community-wide prevention and intervention efforts. [unreadable] [unreadable] This application proposes secondary data analysis of the largest-ever randomized community trial of underage alcohol use to: (1) provide the first estimates of the magnitude of clustering of underage alcohol use based on pair wise odds ratios obtained from a statistical method called alternating logistic regression (ALR), (2) evaluate the role of individual, social and community-level environmental contexts on this clustering, and (3) develop a statistical approach for examining the clustering of subtypes of underage drinkers derived from latent class analysis. The proposed study will capitalize on data from repeated cross-sectional samples of approximately 6800 youth aged 14-20 from 68 US cities surveyed in each of three years (2004, 2006 and 2007) and collected as part of the Enforcing Underage Drinking Laws Randomized Community Trial (EUDL-CT). To the extent to which we can understand what causes clustering of underage alcohol use, we may identify aspects of the environment that are shared within communities and that may help to account for the higher or lower prevalence of underage alcohol use. This information will have importance by disclosing what contextual factors should be targeted in future community-wide prevention and intervention efforts. [unreadable] [unreadable] [unreadable]